ETA for 0.3.0?

[yarikoptic]

• @mvdm might be interested to try on his use case
• would it become available/usable in MATLAB as well?

[johnmbarrett]

I am also interested in when this version will be available. I have an experiment where I want to store a bunch of discrete, homotypic, instantaneous, behaviour-related events, so right now my best bet is to store a dataset of the Events type from ndx-events v0.2.0 inside a pynwb.behavior.BehavioralEvents, but to do so I have to sneak it in by exploiting a weakness in the validation of BehavioralEvents.add_timeseries. Based on the extension proposal doc for ndx_events v0.3.0, it seems like the new version should provide a cleaner way to do this.

[weiglszonja]

@rly can you cut a release for 0.3.0?

[TomekFraczek]

@rly is there a planned release timeline for 0.3.0 and 0.4.0?

We developed an extension (https://github.com/BCM-Neurosurgery/ndx-wearables) that depends on and utilizes the powerful features introduced in ndx-events 0.4.0, and would love to release it to PyPI, but can't do so without a official release of ndx-events.

Thanks!

[rly]

The plan now for ndx-events is to release it as part of the next NWB schema release so that an extension is not needed. The timeline for that unfortunately keeps getting pushed back, but it is a top priority. I plan to have it merged and released by the end of the month. (nwb-schema PR)

I had been hesitating to make a release of ndx-events 0.4.0 because then we would need to support another version of ndx-events when supporting backwards compatibility in PyNWB. But I think it is best to not hold back usage of ndx-events 0.4.0 by the community and just make the 0.4.0 release now. Let me review it this week and make sure it's all set based on what was discussed in the NWBEP001 review, and I'll make that release early next week.

[rly]

ndx-events 0.4.0 is now available on PyPI: https://pypi.org/project/ndx-events/